My Hacktoberfest Open-Source AI Challenge Submission: Touch Grass Project
My Hacktoberfest Open-Source AI Challenge Submission Project Overview Built an open-source AI project that gets people off the screen and into the world. The project leverages open-weight models and local inference to enable outdoor experiences without constant screen dependency. What I Built Open-weight model for [specific task - e.g., plant identification, trail navigation] Local inference :…
In a bid to encourage people to disconnect from their screens and engage with the world around them, a new open-source AI project named Touch Grass has been launched. The project utilizes open-weight models and local inference, allowing users to enjoy outdoor experiences without being tethered to constant screen reliance.
The project focuses on achieving specific tasks like plant identification, trail navigation, or birdwatching. It is designed to run entirely on a user's device, without any need for internet connectivity. This makes it perfect for activities such as hikes, visits to gardens, birding trips, or even joining run clubs.
One of the key advantages of Touch Grass is its emphasis on privacy. All processing occurs locally, meaning no personal data is sent to any external servers. This ensures that users can safely enjoy the outdoors without concerns about data privacy. Furthermore, the project also promotes accessibility. It works seamlessly even in areas with no internet, making it a valuable tool for outdoor enthusiasts in remote locations.
The technical execution of the project is robust, built on frameworks such as PyTorch or TensorFlow, and capable of running on either CPU or GPU. The model size is compact, weighing in at around 1.2 billion parameters, yet it still offers impressive performance.
Deployment is straightforward, whether it's through Docker, a native binary, or even a mobile app. The project requires a few key libraries, all of which are clearly listed in the installation instructions. To use the Touch Grass project, users simply need to install it via pip or follow the provided installation instructions. Once installed, the program can be run using python main.py or the specified run instructions.
The beauty of Touch Grass lies in its outdoor usage. Whether it's identifying plants while hiking, navigating trails, or birdwatching, the project offers a unique blend of technology and nature. A live demonstration of the project can be viewed through a provided link or by watching a video demonstration. The source code is also available on a GitHub repository for those interested in exploring or contributing to the project.
This submission was made as part of the Hacktoberfest Open-Source AI Challenge, specifically during Week 1 under the theme 'Touch Grass'. Every valid submission earns a virtual sticker, contributing towards Hacktoberfest 2026 rewards.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
